Hello — I’m Michael. I design and build conversational AI systems that feel natural, reliable, and genuinely useful.

My work focuses on chat and voice experiences powered by large language models, where interaction design, behavior, and system thinking come together. I care about what happens when AI meets real users — especially when things don’t go perfectly.

I build agentic systems that guide users through tasks, handle ambiguity, recover from errors, and deliver outcomes — not just responses. My work spans conversation flows, system prompts, tool use, fallback behavior, and the UX decisions that make AI products feel dependable.

From real-time voice interaction to multimodal AI learning systems, I design experiences that balance human intuition, system constraints, and performance. I prototype quickly so ideas can be tested in practice and refined through iteration.

This is where I explore, build, and refine AI experiences for real-world environments.

Designing AI that behaves clearly, responds usefully, and recovers gracefully, grounded in real-world support, automation, and interdisciplinary study at Wharton.